Assessing the relation between accountability, performance ... · Assessing the relation between...
-
Upload
phungnguyet -
Category
Documents
-
view
216 -
download
0
Transcript of Assessing the relation between accountability, performance ... · Assessing the relation between...
Assessing the relation between accountability, performance and donation levels:
A view from the Dutch NGO sector
Name: Timothy Nahr
Student number: 11410620
Thesis supervisor: prof. dr. B.G.D. O’Dwyer
Date: 21 August 2017
Word count: 15912
MSc Accountancy & Control, specialization Accountancy
Faculty of Economics and Business, University of Amsterdam
2
Statement of Originality
This document is written by student Timothy Nahr who declares to take full responsibility for
the contents of this document.
I declare that the text and the work presented in this document is original and that no sources
other than those mentioned in the text and its references have been used in creating it.
The Faculty of Economics and Business is responsible solely for the supervision of completion
of the work, not for the contents.
3
Preface
This master thesis marks the end of my master in Accounting and Control at University of
Amsterdam. Back in January 2017, I chose to write my thesis about NGOs because of my
personal interest and prior personal experiences in this sector.
My interest in NGOs and social enterprises has arisen since I was sixteen years old. At the time I
was given the opportunity through an NGO project named ‘Fundashon Bon Intenshon’ to
transfer from Curaçao to the Netherlands to further develop myself in the social and sports field
at the professional soccer school of NEC/Fc Oss. This foundation provides different
scholarships and public benefits to the less advantaged people on the island, so that they can
further develop their talents and maximize their potential. As this opportunity has been so
crucial to my development as a person, my goal has always been to develop sustainable programs
and businesses that could create long term value for society and in particular those in need.
By researching the NGO-donor relationship I felt that I could gain more insight into my field of
interest.
My appreciation goes to the staff of ‘Centraal Bureau Fondsenwerving’ (CBF) for providing me
with the necessary dataset for me to do this research. I also would like to thank ‘Fundashon Bon
Intenshon’ for all the help they have given me throughout the years. Moreover, I would like to
show my gratitude to my thesis supervisor prof. dr. B.G.D. O’Dwyer for challenging me in the
process and providing me with insightful vision regarding this topic throughout his sustainability
lectures. Similarly, I would like to thank my student colleagues for their collegiality throughout
the academic year.
Lastly, a special thank-you goes to my family, host family and close friends who have been there
for me from day one. You have been such a great source of inspiration and motivation and
hopefully one day I can make you proud.
I hope you may read this thesis with the same pleasure as I had writing it.
Your sincerely,
Timothy Nahr
4
Abstract
Purpose: The aim of this thesis is to examine the relation between accountability, performance
and donation levels in the Dutch NGO sector. Hereby to see if a higher level of accountability
would influence NGO performance and whether higher accountability leads to a significant
increase in donations. If not, the NGOs find themselves in a reputation trap for no reason. The
main research question is: What are the effects of accountability on NGO performance and
donation levels?
Design/methodology/approach: To answer the research questions and the sub-questions a
quantitative research is conducted. First the relevant scientific literature is discussed to get an
insight on the different notions. Afterwards, the collected data is statistically tested and the
results make it possible to answer the research question. The data sample consist of Dutch
NGOs with a CBF certification mark, who participated in the Transparency prize in the year
2011-2015.
Findings: Based on the regression results it can be concluded that donors pay more attention to
the efficient usage of program funds than to administration and fundraising costs. Also there is
no evidence found that higher accountability leads to a higher NGO performance. Furthermore,
there is no significant positive relation found between accountability and donation levels. It
could be stated that the Dutch NGO sector lies in a reputation trap for no reason as the donors
don’t seem to be interested in more transparency.
Originality/value: The literature review identified a lack of empirical data available on
accountability within NGO contexts; this study tries to fill that gap.
Limitations: Not all Dutch NGOs participated in this study, only NGOs with a CBF
certification mark. Although transparency is the literal value of accountability, the proxy used to
measure accountability (namely the transparency prize), might not capture all aspects of
accountability.
Keywords: Accountability; Non-governmental organization (NGO); Donation; Performance.
5
Contents
1. Introduction 7
1.1 Research question 9
1.2 Motivation for study 10
1.3 Thesis outline 11
2. Literature review and Hypotheses development 12
2.1 Defining and classifying NGO 12
2.2 Accountability 13
2.3 Accountability in NGO-donor relationship 16
2.4 Effect of performance on donation levels 18
2.5 Dutch nonprofit sector 19
2.6 Hypotheses development 21
3. Research method 23
3.1 Data and sample selection 23
3.2 Measurement of variables 25
3.2.1 Accountability 25
3.2.2 Performance 25
3.2.3 Donation levels 26
3.3 Control variables 26
3.4 Regression models 29
4. Results 31
4.1 Descriptive statistics 31
4.2 Bivariate analysis 35
4.3 Regression analysis 38
5. Conclusion 50
5.1 Summary 50
5.2 Implications 52
5.3 Limitations and recommendations for future research 52
References 54
Appendix 59
6
List of Figures and Tables
List of Tables
Table 1: Classification of NGO activity 13
Table 2: Conceptions of accountability 16
Table 3: The reputation trap 18
Table 4: (In)dependent variables 24
Table 5: (In)dependent variables 28
Table 6: Control variables 29
Table 7: Descriptive statistics of all the variables 34
Table 8: Spearman-Rho matrix for all variables over six-year period 37
Table 9: Regression estimates of accountability on performance (ADMIN) 40
Table 10: Regression estimates of accountability on performance (PROG) 41
Table 11: Regression estimates of accountability on performance (FR) 42
Table 12: Regression estimates of performance on donation levels (ADMIN) 45
Table 13: Regression estimates of performance on donation levels (PROG) 46
Table 14: Regression estimates of performance on donation levels (FR) 47
Table 15: Hypotheses conspectus 48
Table 16: Regression estimates of accountability on donation levels 49
List of Figures
Figure 1: Conceptual model 22
Figure 2: Division of NGOs across sectors 33
Figure 3: Ordinal scores for TP prize 33
7
1. Introduction
Due in part to a series of highly publicized scandals or failures by non-governmental
organizations (NGOs) to live up to public expectation, the confidence level in NGOs has
declined over the past decade (Ebrahim, 2009). Following an increased demand to prioritize
improvement of accountability within the international aid agenda (OECD, 2011) and
transparency emerging as a key aspect of good development (OECD, 2012), this paper examines
the relationship between accountability, performance and donation levels in the Dutch NGO
sector. The objective of this study is to determine whether performance can be improved when
having a higher level of accountability and whether these factors significantly effects the
donation levels (donors are crucial for the survival of NGOs).
Over the last decade there has been a dramatic growth in the number of NGOs, such as those
involved in development aid, in both developed and developing countries. Many NGOs are
prominent in attempts to improve the lives of disadvantage people and have traditionally been
deemed more trustworthy than governments and corporations (O'Dwyer & Boomsma, 2015).
Public funds being spent through NGOs has grown dramatically and the proportion of
development aid going through NGOs has also increased. Because of the large increase in public
funds going through these organizations, NGOs have become subject to much more critical
scrutiny regarding their accountability by the stakeholders (Unerman & O'Dwyer, 2006a). Most
academic studies related to accountability focus on businesses. However, NGO accountability
has gotten increasing attention in recent years (Unerman & O'Dwyer, 2006a).
Associated with this growth has been a growing concern about identifying the achievements of
NGOs. Therefore, performance measurements to asses NGOs effectiveness are becoming more
important. The importance of performance measurements is evident in the literature on the
monitoring and evaluation of the activities in non-profit organizations (NPOs) (Kaplan, 2001;
Beamon & Balcik, 2008). Research by ING (2016) on the NGO-donor relationship in the
Netherlands, indicated that a lack of trust due to slow economic growth after the last financial
crisis, led donation levels to stagnate over a long period of time. Therefore, a higher level of
transparency is needed to regain this trust. O’Dwyer & Unerman (2008) stated that the
competitively driven, donor dominated NGO world, now seeks greater external accountability
8
mechanisms to achieve this level of transparency. While competing against the other NGOs,
transparency is crucial to maintain a share in the tight funding budget.
NGOs are NPOs that exist to provide a public benefit. They can be categorized as an
intermediary between donors (who provide resources, usually in the form of time or money) and
beneficiaries (who are the recipients of the benefit being provided) (Hyndman & McDonnell,
2009). This research focuses on NGOs in the Netherlands, as the Netherlands is in the top 10 of
the most generous countries according to the World Giving Index 2015. It has also been widely
credited with leading international efforts to assess the quality and effectiveness of development
aid. The Dutch government has consistently stressed the need to focus on assessing the results
of development aid efforts, most recently in the context of the UN’ s Millennium Development
Goals (MDGs) (Ruben & Schulpen, 2009).
Accountability involves explaining what you have done and taking responsibility for your actions
(Unerman & O'Dwyer, 2006b). It thus follows that NGO performance must be judged from the
perspectives of those who affect or are affected by the organization's behavior (Mitchell, Bradley,
& Wood, 1997). Most importantly, these stakeholders are those who contribute to the NGOs
survival. Moreover, there are cases of misused of funds by NGOs, for example, in a
humanitarian aid NGO in Syria, which was founded bribing ISIS officials in order to continue
working in territory under the Islamic State’s control. These scandals would trigger calls from
donors, demanding larger accountability and transparency from the whole NGO sector. A lack
of accountability might not only be harmful for the involved NGOs, but also for the sector as a
whole, as it can lead to a reduced public support for all NGOs (Steinberg, 2006).
The Netherlands as one of the leading giving countries, has paid a lot of attention to
accountability in recent years. An organization who focusses on the transparency in the Dutch
NGO sector is the ‘Stichting Centraal Bureau Fondsenwerving’ (CBF). CBF monitors the NGO
sector and provide certification marks based on specific criteria. With this certification mark,
NGOs can signal their good behavior to the market. Additionally, the website ‘Goede doelen
monitor’, founded in the Netherlands, provides information about NGOs to help donors find an
organization that best suits them.
9
1.1 Research question
As stated earlier, scandals and the misuse of funds by NGOs would trigger calls from donors,
demanding larger accountability and transparency. Research by ING (2016) showed that
donations are often based on the reputation and thus the performance of an NGO. As these
elements have become crucial for the donation levels, NGO managers have incentives to
perform well. Therefore, understanding NGO accountability mechanisms and the relationship
with donation levels and performance measurement is important for NGO managers.
NGOs need donors to survive and donors need NGOs to implement shared policy goals (Gent,
Crescenzi, Menninga, & Reid, 2014). According to Gent et al. (2014), as much as donors want to
provide NGOs with resources to help them achieve their missions, they are also concerned with
the quality of their investments. Due to the uncertainties regarding quality of the individual
NGOs, donors want guarantees that the resources they provide are being used effectively. As
NGOs know that donors try to invest in NGOs that have demonstrated their quality, NGO
managers have an incentive to focus their efforts on short term achievements that are easily
attributable to the NGO. Even when the activities differ from the overall mission of the NGO,
the NGO may feel to divert resources to consistent and public demonstrations of tangible assets
in order to preserve funding. Herein lies a reputation trap (Gent et al., 2014). Some studies have
acknowledged that implementing inappropriate accountability mechanisms can damage, instead
of enhance the benefits NGOs seek through their projects (Ebrahim, 2003a; Ebrahim, 2003b;
Unerman & O’Dwyer, 2006b).
Accordingly, the emerging dominance of upward accountability to donors at the possible
expense of more accountability to a broader range of stakeholders, especially beneficiaries, has
created concerns that NGOs’ accountability priorities are being distorted (Ebrahim, 2005;
Ebrahim, 2003a; O’Dwyer & Unerman, 2007). But is it beneficial for NGOs to settle for non-
durable outcomes? And is the effect of accountability on the donation levels significant?
The aim of this study is to analyze whether a higher level of accountability would influence the
level of performance of an NGO and whether this higher accountability leads to a significant
increase in donations. If not, would the NGOs be better off freeing themselves from the
reputation trap and focusing on durable impact? Based on these aims, the following research
question was formulated:
10
RQ: What are the effects of accountability on NGO performance and donation levels?
1.2 Motivation for study
This research definitely has social and practical relevance. It aims to contribute to prior research
on NGO accountability by filling the gap caused by limited evidence about the relationship
between accountability, performance, and donation levels for NGOs. This research directly
responds to calls to examine the emergence of accountability within development NGO contexts
(see, Ebrahim, 2009; O’ Dwyer & Unerman, 2008). Secondly, prior studies also focused on
accountability and performance evaluation in NGO’s separately. Third, as funds are a key point
of survival, a higher performance should lead to a higher donation levels. As the actions of these
NGOs can have a substantial effect on the lives of others, both directly and indirectly, by
studying both types of relationship in the same research, a better understanding is obtained of
the relationship between accountability, performance, and donation levels. Lastly, the new Dutch
financing scheme (MFS 2) covering the funding period of 2011-2015, could have had an effect
on how accountability is viewed in society and how the NGOs implement it in practice.
Therefore, examining the relationship between accountability, performance and donations level
for the period 2010-2015 will contribute to the NGO accountability literature.
This study also provides societal contributions. The results could be useful to donors, policy-
makers, NGO managers and other stakeholders by providing new insights about whether more
accountability leads to higher performance and donation levels. Following an increase demand to
prioritize improvement of accountability within the international aid agenda (OECD, 2011), the
results could help policy-makers further standardize the accountability framework or help NGO
managers when making strategic decisions.
Performance is hard to measure in the non-profit sector. As stated in O’Dwyer & Unerman
(2008), some managers feel that the quantitative metrics to determine performance don’t always
fairly represent their fieldwork. Miller (2002) also states that the non-profit sector lacks a
commonly accepted performance indicator. The aspect that will be looked at in the rest of the
paper is efficiency. It is how efficient the NGOs are with the money they are receiving and
demonstrating that they are investing the money in durable outcomes.
11
1.3 Thesis outline
This study is structured as follows. First, the literature review and the hypothesis development
will be described. Chapter three will describe the data methodology and the used sample. The
fourth chapter will contain an analysis of the research results. The final section summarizes the
findings of the research and provides suggestions for future research.
12
2. Literature review and Hypotheses development
The research question: ‘‘What are the effects of accountability on NGO performance and donation levels?’’
entails three main concepts, namely NGOs, their performance and the NGO-donor relationship.
This chapter reviews the literature written on NGOs and their performance measurements. Also
the NGO-donor relationship will be discussed. Since this paper focuses on Dutch NGOs, the
charitable sector in the Netherlands will be discussed. This section will be concluded with the
hypotheses development.
2.1 Defining and classifying NGO
There is a notion that NGOs differ from businesses because they pursue principled beliefs, while
businesses pursue material interests (Sell & Prakash, 2004). O'Dwyer & Unerman (2007) state
that there is little consensus on how to define and classify NGOs due to their varying in size,
topical coverage and scope. As such there are different definitions of NGOs in the literature. To
understand the actions of NGOs and businesses it is important to first distinguish between both
and provide a definition of NGO.
According to Hansmann (1980), an NGO is an organization that is barred from distributing its
net earnings. Thus, an NGO is characterized by its constraint in distributing its net earnings to
any individuals who exercise control over it, such as members, officers, directors or trustees. An
NGO is distinguished from a for-profit business primarily by the absence of a right of the
control group or ownership to share in the profits. But a price could be expected to be paid in
incentives due to the elimination of the profit motives. NGOs do succeed in distributing some
of their profits through inflated salaries and various other perquisites granted to employees
(Hansmann, 1980).
Edwards (2000) defines NGOs as ‘‘a subset of civic organization, defined by the fact that they
are formally registered with government, receive a significant proportion of their income from
voluntary contributions and are governed by a board of trustees rather than the elected
representatives of a constituency.’’ But Edwards’ choice of ‘registration’ might be flawed. As
there are forms of registration that might apply to charities, grant-receiving bodies, community
based enterprises and so on, which are not necessarily considered NGOs and/or not registered
13
in this manner (Gray, Bebbington, & Collison, 2006). In this paper, an NGO refers to an NPO
that exist to provide a public benefit. Hereby, they form an intermediary between donors (who
provide resources, usually in the form of time and money) and beneficiaries (who ultimately
benefit from the donations) (Hyndman & McDonnel, 2009).
Furthermore, there are different types of NGOs. The main factor that distinguishes between
types of NGOs is the kind of activity the NGO engages in (Van Tulder & Van Der Zwart,
2006). Based on the activity NGOs engages in, they can be classified into the following four
categories: 1) social welfare, 2) international aid, 3) health, and 4) environmental movement
(Centraal Bureau Fondsenwerving, 2017). Each field has organizations that focus on one specific
issue, like human rights, victim aid, public health or environmental campaigns (van Tulder & van
der Zwart, 2006). Table 1 provides an overview of the types of NGOs based on a specific
activity.
Table 1
Classification of NGO activity
(Centraal Bureau Fondsenwerving, 2017)
2.2 Accountability
Many NGOs are prominent in attempting to provide a public benefit and improve the lives of
disadvantaged people. They have also traditionally been deemed more trustworthy and able to
provide these services than governments (O'Dwyer & Boomsma, 2015). But concerns about
accountability in NGOs have increased in recent years, due in part to some highly publicized
scandals and NGOs failing to meet certain expectations, deteriorating the public confidence in
the charitable sector (Ebrahim, 2003a). As a response, several transnational NGOs have
acknowledged their accountability to a range of constituencies by developing a series of
accountability mechanisms (Ebrahim, 2009) and signing up to accountability charters (Schmitz,
14
Raggo, & Bruno-van Vijfeijken, 2012). Dealing with increased critical scrutiny by the public is
important for these NGOs due to the concern that donors can punish these NGOs and cut their
funds (Ebrahim, 2003b).
As an abstract and complex concept, the term accountability has lacked a clear definition
(Ebrahim, 2003b). There are several definitions of accountability in the literature. Edwards &
Hulme (1996b) define accountability as “the means by which individuals and organizations
report to a recognized authority and are held responsible for their actions”. In their study of
NGO accountability, Unerman & O’Dwyer (2006b) describe accountability as ‘‘a process to
explain and take responsibility for your actions.’’ Unerman & O’Dywer (2006b) argue that ‘‘the
main purpose of accountability is to provide mechanisms through which all those affected by an
organization’s or person’s actions can demand an account from the person or the managers of
that organization regarding how and why they acted in that manner.’’ Accountability is also
defined as ‘‘the means through which individuals and organizations are held externally to account
for their actions and as the means by which they take internal responsibility for continuously
shaping and scrutinizing organizational mission, goals, and performance’’ (Ebrahim, 2003b).
Not only the socially constructed nature of the term accountability makes it hard to define, but
organizations also often face multiple accountabilities that change over time (Ebrahim, 2003b).
Although the definitions may differ, the concept of accountability in general covers the aspect of
trust. This trust is connected to the timely availability of reliable information, which is essential
for the performance measurement and monitoring of NGOs by beneficiaries, donors and
governments. As donors are usually physically removed from the site of the NGO activities, they
are reliant on the statements provided by the NGOs to assess their performance (Burger &
Owens, 2010). Transparency is a key issue in the NGO sector due to information asymmetries
(Burger & Owens, 2010). Access to information is a key characteristic of accountability, since all
accountability relies on relevant and timely information (Cameron, 2004). According to Koppel
(2005), transparency is an important instrument for performance measurement and is a key
requirement for all other dimensions of accountability (see Table 2).
As transparency is the literal value of accountability (Koppel, 2005), this study uses the
transparency prize as a measure of accountability. The criteria used for the transparency prize
(see Appendix 2) are aligned with the concept of information availability and are therefore
related to the concept of transparency, one of the aspects of accountability.
15
NGOs are accountable to a range of stakeholders (O’Dwyer & Unerman, 2008). According to
Ebrahim (2003a), “hierarchical accountability is narrowly focused, short-term in orientation and
favors accountability to the more influential stakeholders (upward accountability)’’. Hierarchical
accountability is perceived to be a form of accountability of external oversight and control and
encourages the rationalization of actions. NGOs must constantly try to quantify their
performances to be assessed by this group of stakeholders (O’Dwyer & Unerman, 2008). To
survive financially, NGOs become frustratingly hobbled by the search for tangible results to
maintain their reputation (Gent et al., 2014). Hence, it is often claimed that this type of
accountability can have negative effects on the effectiveness of an NGO (Ebrahim, 2005).
A much broader form of accountability, where the impact that the organization can have on
other organizations, clients and their environment is taken into account, is called holistic
accountability or downward accountability (O’Dwyer & Unerman, 2008). Within holistic
accountability, in addition to the key stakeholders recognized under the hierarchical
accountability, an NGO is accountable to the group on whose behalf the NGO advocates (i.e.,
beneficiaries), along with the people, communities or regions indirectly or directly impacted by
the NGOs activities (Ebrahim, 2003a) as cited in O’dwyer & Unerman (2008). Since funders are
key to the survival of an NGO and because beneficiaries are usually passive (O'leary, 2017) this
study focuses on upward accountability. Therefore, this study further examines the NGO-donor
relationship.
Various theories have been used to analyze why NGOs engage in accountability practices.
Related theories are the stakeholder theory and legitimacy theory. In stakeholder theory, the role
of management is seen as achieving a balance between the interests of all stakeholders. The only
way to ensure that the firm can survive is by maintaining the different stakeholders happy. A
business has multiple contracts with different stakeholders. Without the participation of the
primary stakeholders, an organization cannot exist (Deegan & Blomquist, 2006). So being
accountable keeps the stakeholders happy, because of the responsible image the firm is
portraying, and so the firm can survive.
Another reason for NGOs to engage in accountability practices is because they have a social
contract with society and are deemed to behave in a way that is proper within the community in
which they operate (Deegan & Blomquist, 2006). Thus, the actions of the NGO managers are
aligned with the social expectations of the NGO. By this proper behavior, the NGOs receive
16
the legitimacy to exist (Deegan & Blomquist, 2006). Providing a legitimating symbol can sustain
the existence of the NGO through support by donors (Deegan & Blomquist, 2006).
Table 2
Conceptions of accountability
(Koppel, 2005)
2.3 Accountability in NGO-donor relationship
NGOs often work on temporary and even counterproductive accomplishments. In that
situation, there is incongruence between the long-term goals of the NGO and the outcomes. But
why do these NGOs settle for non-durable outcomes? One explanation is a strategic behavior
between these altruistic organizations and their donors (Gent et al., 2014).
The relationship between an NGO and the donors is a classic principal-agent relationship (Gent
et al., 2014). The agency theory is one in which a group or actor (principal) attempts to have its
agendas carried out by another group or actor (agent) (Ebrahim, 2003a). But sometimes, the
desires or goals of the agent are not the same as those of the principal and it is difficult or
expensive for the principal to monitor the agent’s actions (Shankman, 1999). In this case, there is
information asymmetry. This is known as the agency problem.
NGOs face constant pressure to please donors to maintain financial stability and continue to
pursue their policy goals. On the other hand, donors are faced with a large sector of various
NGOs and must decide in which NGO to invest. These donors are skeptical and do not want to
17
make a bad investment in a low-quality NGO (Gent et al., 2014). As the donors cannot separate
the bad NGOs from the good ones, this creates a situation of uncertainty, known as the lemon
problem.
According to Gent et al. (2014), these sources of uncertainty put donors in a situation where they
might not be able to maximize their investment in an NGO. If donors cannot assess the level of
quality, they might end up investing their resources in a low-quality NGO. This agency problem
is known as adverse selection and emerges when principals cannot select the type of NGO that
they would choose if they had complete information (Gent et al., 2014). Additionally, principals
are usually physically not on site to monitor the behavior of the agents. This problem, which
results from the inability of the principals to capture hidden behavior, is known as moral hazard
(Gent et al., 2014).
To resolve the information asymmetry and ensure that the donor can distinguish low and high-
quality organizations, donors look for measurable performance indicators by the NGOs.
Knowing the objectives of the NGO is not enough for the donor, being able to monitor and
quantify the extent to which the NGO has achieved its goals is also necessary (Gent et al., 2014).
The mechanism that naturally emerges to address the principal-agent problems between donors
and NGOs is reputation (Gent et al., 2014). To maintain their funding, NGOs are tempted to
aim at immediate policy successes. Donors’ concerns about the resources they have invested and
demands to assess the quality of the NGO, discourage NGOs from investing in durable
outcomes. Instead, the NGO will choose the option that would signal to the donors, in the short
term, that they can produce desirable outcomes (Gent et al., 2014). Unerman & O’Dwyer
(2006b) state that implementing inappropriate accountability mechanisms can damage rather
than enhance the benefits NGOs seek to achieve. Even when such activities differ from the
overall mission of the NGO, the NGO may feel the need to divert resources to consistent and
public demonstrations of tangible assets to preserve funding. Herein lies a reputation trap (Gent
et al., 2014).
Table 3 summarizes the NGO-donor relationship and its influence on durable policy outcomes.
An NGO can pursue activities that are likely to be attributable to the mission of the NGO or
those that are likely not to be attributable. Pursuing non-attributable outcomes would not help
fix the problem of information asymmetry, as these outcomes do not provide relevant
18
information needed by the donors to assess the NGO’s performance (Gent et al., 2014). On the
other hand, accountability mechanisms motivate NGOs to pursue attributable outcomes to
signal quality behavior. In an ideal situation, an NGO would pursue activities that would help
them achieve their mission. However, the attributable strategies available to NGOs lead to policy
outcomes that are not durable (Gent et al., 2014). Gent et al. (2014) argue that NGOs and
donors may find themselves in this upper left-hand corner of Table 3 as a result of the
reputation trap.
Table 3
The reputation trap
(Gent et al., 2014)
2.4 Effect of performance on donation levels
As mentioned earlier, there is a lack of a common performance indicator in the non-profit
sector. According to Speckbacher (2003), performance measurement in NGOs is difficult, as
there is no primary interest group that is invariably and clearly defined. NGOs are built around a
specific mission, which is not easy to measure, and they serve different beneficiaries, which
might have different goals. Therefore, it seems that performance measurement tools of the for-
profit sector are not transferable to the non-profit sector (Speckbacher, 2003).
The relationship between performance in NGOs and donation levels have been researched and
the results have been mixed. Tinkelman & Mankaney (2007) investigate the effect of
administration efficiency (ADMIN), i.e., administrative costs divided by total costs, on charity
donation levels. They find that if the organization is primarily dependent on donors for survival
19
and reports nontrivial administrative and fundraising costs, donors use administrative ratios for
their donation decisions. In such cases, managers need to account how their spending is in line
with the organizational goals. Overall, there is a significant negative association between ADMIN
and donation levels found. Conversely, when the organization primarily depends on program
fees or other sources of revenue, the accountability of administrative spending might be ignored
by donors. While Greenlee & Brown (1999) find that ADMIN has a significant negative
association with donation levels, Frumkin & Kim (2001) find no significant relationship between
ADMIN and donation levels.
Jacobs & Marudas (2009) examine the relationship between performance efficiency and donation
levels by building a testing model with the measures ADMIN and PRICE, where PRICE is the
total expenses divided by the program expenses. Using archival data from a sample of US NPOs,
they found that ADMIN and PRICE has a significant negative effect on donations.
Given the mixed results on the relationship between NGO performance and donations, it is
difficult to draw conclusions about the nature of the relationship. This research investigates the
relationship between the effect of NGO performance on donation levels in Dutch NGOs. As
performance is so hard to measure in NGOs, efficiency is used as a stand-in for performance in
this study.
2.5 Dutch nonprofit sector
In the mid-1990s, the Dutch nonprofit sector emerged as among the largest in the world
(Brandsen & Pape, 2015). Besides the NGOs, volunteers and the government, other third parties
are active in the nonprofit sector, which helps stabilize the sector. The biggest is the ‘Stichting
Centraal Bureau Fondsenwerving’ (CBF) (translated: Dutch Central Fundraising Agency). The
CBF is the supervisor of the nonprofit sector. It was founded in 1925 by Dutch municipalities
and NGOs due to the importance of supervision over fundraising activities in the NGO sector.
The CBF describes its mission as “securing the public confidence in the sector and further
development of the nonprofit sector” (CBF, 2016).
The CBF is also an independent accrediting agency. The CBF certification mark assures the
public that the accredited NGO can be trusted and that their donations will be handled
20
responsibly. The NGOs are judged based on various criteria, such as: good governance, policies,
fundraising, use of funds and their accountability towards their stakeholders (CBF, 2016).
In 2015, the ‘Nederland Filantropieland’ (FI), ‘Goede Doelen Nederland’ and CBF developed
the ‘Erkenningsregeling’ (ER) (translated: qualification system) for nonprofit organizations (CBF,
2016). In 2016, the ER replaced the existing CBF certification mark, ‘RfB keur’ and the
‘Keurmerk Goede Doelen’. The goal of the ER is to make the nonprofit sector more
professional, trustworthy and transparent (CBF, 2016). The new guideline means that there is
one new certification mark in the sector. By combining the certification marks (into the ER), it is
easier for the donors to assess the quality of the different organizations, as the NGOs can now
show that they meet all quality requirements through one seal. The ER also takes the size of the
organization into account (the larger the organization, the stricter the norms).
Another third party active in the Dutch nonprofit sector is ‘Goede Doelen Nederland’ (GDN)
(translated: charity Netherlands), a branch organization that helps create publicly recognizable
and respected guidelines within which the sector operates and which supports the functioning of
NPOs. Furthermore, they support their members in conducting efficient business operations and
provide a broad platform for knowledge exchange.
The audit firm PWC is also a party in the Dutch nonprofit sector, as they hand out the yearly
transparency prize. In 2004, they created this award to encourage NGOs and social enterprises
to be more accountable and transparent in their annual reports to the public. PWC collects
annual reports on the participating NGOs and grade them based on several criteria (see
Appendix 2). Based on these criteria, each NGO is awarded a transparency score. As the prize
encourages NGOs to be more transparent and show a high level of accountability for their
actions, this can enhance the reputation of the whole sector (Transparant Prijs, 2015). The
transparency score will be used in this study as a proxy for the level of accountability (see Section
3.1).
As mentioned before, the Netherlands is one of the top countries regarding philanthropic
activities (ING, 2016). Dutch NGOs collected €4,1 billion in total revenues in 2015. About €3
billion came from fundraising activities and government support. The other €1,1 billion were
collected by third-party action and revenue from own investments (e.g., interest received). Of
21
this €4.1 billion, €3,5 billion was ultimately spent on achieving the primarily organizational goal
(CBF, 2017).
2.6 Hypotheses development
This section presents the hypotheses tested in this research, based on the aforementioned
literature and theories. In Figure 1, an overview of the interdependencies between the main
variables is given in the form of a conceptual model.
As analyzed in the previous section, NGOs have various motives to engage in accountability
practices. The stakeholder theory stresses the importance of the role of management in achieving
a balance between the interests of all stakeholders. An NGO has multiple social contracts with
society and needs to keep the primary stakeholders content to survive (Deegan & Blomquist,
2006).
The legitimacy theory stresses the importance of behaving properly within the community in
which the NGO operates (Deegan & Blomquist, 2006). For this proper behavior, it receives
legitimacy to exist on the market (Deegan & Blomquist, 2006). This is also necessary for the
organization to survive (Cho, Guidry, Hageman, & Patten, 2012).
As derived from the stakeholder theory and legitimacy theory, being accountable keeps the
stakeholders happy because of the responsible image the NGO is portraying, which helps the
NGO survive. Miller (2002) explains how accountability leads to a better performance. He
argues that the board of the NGOs create the performance measurements for the management
based on the performance expectations of the various stakeholders. Using the performance
measures, the board can evaluate the management’s performance and monitor whether their
activities are in line with the organizational goal and thus improve performance. Additionally, the
board is held accountable to the various stakeholders and communicates whether management
performance is in line with their expectations.
However, Buchneit & Parsons (2006) found that greater accountability doesn’t necessarily
improve the performance in NGOs. They argue that although stakeholders, such as donors,
22
prefer to know where their resources are going, accountability practices do not actually improve
the performance of the organization. Due to the mixed results, this study will test the following
hypothesis: H1: The level of accountability is positively related with the level of performance
The research by ING (2016) states that donors base their donation decisions on the reputation
of the NGO. Whereas Greenlee & Brown (1999) find that ADMIN has a significant negative
association with donation levels, Frumkin & Kim (2001) find no significant relationship.
Furthermore, research by Jacobs & Marudas (2009) found a link between performance and
donation levels in the US. They found a negative relation between the performance measures
ADMIN and PRICE and donation levels. This study tests this link for the Dutch NGOs, leading
to the following hypothesis: H2: Higher performance is positively related to donation levels
As accountability is a form of good governance, providing relevant and reliable information can
mitigate the agency problem and ensure selection of high-quality NGOs by donors (Gent et al.,
2014). The research of ING (2016) states that an increase in the level of transparency through a
report or other forms of media coverage enhances donor trust and leads to a stronger relation
with the donors. It is expected that a higher level of trust translates into a higher donation levels.
The corresponding hypothesis is: H3: A higher level of accountability is positively related with donation
levels
Figure 1
Conceptual model
Accountability
Performance Donation level
H2 (+)
23
3. Research method
This section elaborates on the methodology employed to test the hypotheses developed in
Chapter 2.
3.1 Data and sample selection
To test the relationship between accountability, performance and donation levels in the NGO
sector, data has been collected using a quantitative method. Quantitative research was chosen as
it is an effective tool to test the empirical evidence of predetermined hypothesis (Ahrens &
Chapman, 2006).
The NGO accounting performance data was obtained from the CBF database. The CBF
database provides access to NGO accounting data and information from NGOs in the
Netherlands. They have the largest database, being the leading accreditation agency in the
Netherlands in this sector. The CBF keeps a register of public data, e.g., annual reports, that are
handed in by the NGOs themselves or are collected by CBF employees. As the financial
information of the NGOs is audited by a credited audit firm, the reliability of the collected
information is ensured. Any missing data is complemented with data from the annual reports
and from the NGOs’ websites.
The CBF data is complemented by the information from the transparency prize. This data is
obtained from transparency prize archive from the audit firm PWC. PWC has handed out this
award for NGOs since 2004. The transparency prize measures how transparent the participating
NGOs are through their annual reports, based on a list of criteria (see Appendix I). Thereafter,
each NGO receives a transparency score (TP). Big NGOs are ones that have more than €0,5
million in yearly revenues and small NGOs those with less than €0,5 million. The criteria for the
big and small NGOs only differ slightly in the categories governance and communication (see
Appendix 1).
The sample is the Dutch NGOs that participated in the transparency prize in 2011, 2012, 2013,
2014 and 2015. Thus, the financial data used for these organizations consists of the years 2010
up to and including 2015.
24
The transparency score is used as proxy for NGO accountability. Of the organizations that
participated in the transparency prize, only the NGOs with a CBF certification mark (i.e., ER)
are used (reference date March 22nd, 2017). After matching transparency prize data with financial
information from CBF database, an initial sample of 216 NGOs was obtained. Removing
organizations that do not have a CBF certification mark resulted in a sample of 162 Dutch
NGOs out of 1492 registered NGOs in the CBF database (reference date March 22nd, 2017).
After controlling for outliers, the initial sample of 162 NGOs was cut down to 150 (see Section
4.1). This is a coverage of 10% of the registered NGOs at the CBF. It must be noted that the
CBF has more NGOs registered in their database than NGOs that holds a CBF certification
mark. For an overview of the sample selection, see Table 4.
The transparency prize was chosen, as a reliable proxy for accountability was needed to
investigate accountability. The reliability of this proxy is further strengthened by the reputation
of the big-4 audit firm PWC in assessing organizational performance. With the experience and
credibility of PWC in the audit field, this proxy is found in the TP. Furthermore, as TP data was
only available for 2011–2015, only participating organizations in these years were used in the
study. Taking only NGOs that have a CBF certification mark ensures that they all have some
common characteristics. As mentioned earlier, as the criteria for grading big and small NGOs
differ slightly and size is used as a control variable in this study (see Section 3.3), no further
distinction has been made between these two groups in this study. The above information is
considered sufficient to determine the level of accountability and performance and the control
variables.
Table 4 Sample selection
Original number of organizations
Observations
1,492
Less all organizations that did not participate in TP (1,276)
Less all organizations with no CBF certification mark
Less outliers
(54)
(12)
Remaining organizations in sample 150
25
3.2 Measurement of variables
3.2.1 Accountability
The independent variable used in regression models 1.1, 1.2, 1.3 and 3 is accountability. As
mentioned earlier, the accountability scores were obtained from the transparency prize archive.
The TP is used as the proxy for accountability per NGO. It is a score from 0 to 10 and measures
the extent to which an NGO communicates relevant and reliable information to the
stakeholders. The TP is divided into four categories: ‘de kopgroep’ (NGOs with a score of 7.5-
10), ‘de achtervolgers’ (NGOs with a score of 7.0-7.5), ‘het peloton’ (NGOs with a score of 5.5-
7.0) and ‘de achterblijvers’ (NGOs with a score <5.5).
As the absolute scores are usually only available for the top ten NGOs, the ordinal scores are
used. To test hypotheses one and three, TP is split into four categories. If the NGO score is
between 7.5 and 10, the TP takes a value of 1. If the NGO score is between 7.0 and 7.5, the TP
takes a value of 2. If the NGO score is between 5.5 and 7.0, the TP takes a value of 3, while the
TP takes a value of 4 if the NGO score is <5.5. For a further overview of the variables, see
Table 5.
3.2.2 Performance
Performance was a dependent variable in regressions 1.1, 1.2, 1.3 and an independent variable in
regression 2. As mentioned earlier, according to Miller (2002) the non-profit sector lacks a
commonly accepted indicator for performance. Therefore, efficiency is used as a proxy for
performance. The three proxies for performance indicators that will be used are: administrative
efficiency (ADMIN), program expenditure (PROG) and fundraising revenue (FR). All these
variables are frequently used in the literature to analyze management performance.
Jacobs & Marudas (2009) use administrative efficiency to measure organizational performance,
while Tinkelman & Mankaney (2007) use administrative efficiency and fundraising results to
analyze the relation between performance and donation levels. Program expenditure is also a
commonly used proxy for performance measurement. Aggarwal et al. (2011) use program
26
spending to measure CEO performance, while Krishnan et al. (2006) use fundraising and
program expenditure to test expense misreporting in NPOs.
ADMIN measures the administration costs as a percentage of the total costs. High
administration expenses is seen as inefficient, therefore, the lower this ratio, the better the NGO
performance. PROG measures the percentage of revenue spent on the programs that help the
organization reach its goals. The more funds are used to achieve the mission the better,
therefore, the higher this ratio, the better the performance. Finally, FR measures the
organization’s costs to raise €1. It is expected that a higher amount spent on fundraising activities
should lead to a higher donation levels (Posnett & Sandler, 1989). But using less funds to raise €1
is the more efficient, therefore, the lower this ratio, the better the performance.
3.2.3 Donation levels
Regressions 2 and 3 used the level of donations (DON_TOT) as the dependent variable.
DON_TOT is the total amount of private donations, e.g., from individuals and companies, and
institutional donations, e.g., government subsidies.
3.3 Control variables
Some control variables were used to test for other possible explanatory variables in the
regression models. Since the activity an NGO is engaged in can influence the performance or the
tendency to be transparent, Sector (INT_AID, HEALTH, ENVIRON, SOCIAL) is taken as a
control variable.
Another control variable is Size. Kahler & Sargeant (2002) found an inverse relation between
ADMIN and size. Bigger organizations can have advantages that smaller organizations do not
have. One factor is the extent to which NGOs can generate attention from prospective donors
(Bekkers & Wiepking, 2011). Size is measured as the total assets on a log scale.
27
Organizational age (AGE) is included as a control variable, as experience may affect NGO
performance. Experience might be attractive to donors, who might reason that for an NGO to
have been around for so long, it must be high quality (Trussel & Parsons, 2008). Having more
experience and learning from your mistakes might lead to positive development in your
performance. According to Weisbrood & Dominguez (1986), there is a positive relation between
goodwill and organizational age. A high level of goodwill means that journalists write positively
about the NGO and that prospective donors are familiar with the brand. As such, the high
quality translates in the organizational age, as the NGO showed that it has survived throughout
the years.
The control variable Transparency price (TPP) is also used, as the TPP suggests that an NGO who
has won the transparency prize in the year of observation, has a high level of transparency and is
functioning well. Winning the award can have a positive effect on the brand. Accordingly, this
could attract more donors. TPP is measured by two dummy variables: the score ‘1’ is given to an
NGO that has won the TPP in the year of observation and ‘0’ to an NGO that has not won the
TPP in the year of observation.
According to Yetman & Yetman (2011), being audited by a big-4 audit firm will impact an
organization, as it leads to more accurate expense reporting. Therefore, the use of a big-4 audit
firm (BIG4) is taken into account. BIG4 is measured by two dummy variables: the score ‘1’ is
given to an NGO that is audited by a big-4 audit firm and ‘0’ to an NGO that is not.
According to Biddle, Hillary & Verdi (2009), managers might have the incentive to relax budgets
and other controls that prevent misbehavior, in case of a large cash balance. As there are enough
funds available, the managers may direct resources to programs or items that do not directly
benefit the organizational goal. Additionally, donors may feel that agency costs could arise from
a high cash balance and adjust their donation behavior accordingly. Cash is measured as the ratio
of cash and cash equivalents to total assets.
The Dutch SHO (Samenwerkende Hulporganisaties) is an organization that unites Dutch NGOs
to respond collaboratively to exceptional disasters. In case of a disaster, the member NGOs join
forces and use the well-known bank account GIRO555 to collect donations for all the member
NGOs (Samenwerkende Hulporganisaties, 2017). NGOs that are part of the SHO may benefit
from the organization’s goodwill and fundraising efforts and therefore receive more donations.
28
SHO is measured by two dummy variables: the score ‘1’ is given to an NGO that is an SHO
member and ‘0’ to an NGO that is not.
Further characteristics of the director, such as gender (DGENDER), may influence performance
and accountability and thus donation levels. Also, a dummy variable for the variable international
parent (INT) is taken into account. Having an INT includes ongoing development and
performance assessment as an organization. It may influence performance and accountability and
thus donation levels. For a further overview of the variables, see Table 6.
Table 5
(In)dependent variables
Variable Variable label Expectation Definition
Accountability
TP
+
Score on the Transparency Prize (TP), 1=score 7.5-
10, 2= score 7.0-7.5, 3= score 5.5-7.0, 4= score
<5.5
Performance Admin - Percentage administration cost of total costs
PROG + Percentage revenue that is spend on organizational
goals
FR - Percentage funding costs to attract fund revenue
Donations DON_TOT NA Total donations composed of private donations +
institutional donations
Notes: “+”, Positive relationship between dependent and independent variable predicted;
“–”, negative relationship between dependent and independent variable predicted.
29
Table 6 Control variables
Variable Variable label Data type Definition
Sector
INT_AID
Dummy
A NGO falls into one category: 1=International aid,
2=health, 3=social welfare, 4=environment
HEALTH Dummy
ENVIRON Dummy
SOCIAL Dummy
Size SIZE Ratio Logged total assets of the organization
Age organization AGE Ratio The age of the organization in years
Transparency prize TPP Dummy Has NGO won transparency prize in the year of
observation? Yes (=1), No (=0)
Auditor BIG4 Dummy Is the NGO audited by a big-4 audit firm?
Yes (=1), No (=0)
Cash CASH Ratio Ratio of cash and cash equivalents to total assets
SHO SHO Dummy Is the NGO a member of SHO?
Yes (=1), No (=0)
Gender executive DGENDER Dummy Gender of the director. 1= Male, 0= Female
International parent INT Dummy Does the NGO have an international parent?
Yes (=1), No (=0)
3.4 Regression models
This paper uses seven models to test the three hypotheses from Chapter 2. The first three
regression models (1.1, 1.2 and 1.3) were used to test hypothesis 1, the influence of
accountability on performance. Accountability was tested with a time lag of 1 year, as can be seen
in the model below. A time lag of 1 year was used because the transparency prize is awarded
after the closing balance date, therefore, the effect of the latter can be noted, if applicable, in a
future period.
(1.1) ADMIN = β0 + β1 Ti,t-1 + d1-4 SECTOR + d5 SIZE + d6 AGE + d7 TPP +
d8 BIG4 + d9 CASH + d10 SHO + d11 DGENDER + d11 INT + ε
(1.2) PROG = β0 + β1 Ti,t-1 + d1-4 SECTOR + d5 SIZE + d6 AGE + d7 TPP +
d8 BIG4 + d9 CASH + d10 SHO + d11 DGENDER + d11 INT + ε
30
(1.3) FR = β0 + β1 Ti,t-1 + d1-4 SECTOR + d5 SIZE + d6 AGE + d7 TPP +
d8 BIG4 + d9 CASH + d10 SHO + d11 DGENDER + d11 INT + ε
Hypothesis 1 is supported if β1 (ADMIN) is negative and significant, β1 (PROG) is positive and
significant and β1 (FR) is negative and significant.
The following regressions tested the influence of performance on donation levels (hypothesis 2).
(2.1) DON_TOT = β0 + β2 ADMIN + d1-4 SECTOR + d5 SIZE + d6 AGE + d7
TPP + d8 BIG4 + d9 CASH + d10 SHO + d11 DGENDER + d11 INT + ε
(2.2) DON_TOT = β0 + β3 PROG + d1-4 SECTOR + d5 SIZE + d6 AGE + d7
TPP + d8 BIG4 + d9 CASH + d10 SHO + d11 DGENDER + d11 INT + ε
(2.3) DON_TOT = β0 + β4 FR + d1-4 SECTOR + d5 SIZE + d6 AGE + d7 TPP +
d8 BIG4 + d9 CASH + d10 SHO + d11 DGENDER + d11 INT + ε
Hypothesis 2 is supported if β2 and β4 are negative and significant and β3 is positive and
significant.
The last regression has the same structure as the previous models. Regression 3 tested the
relation between accountability and donation levels. Accountability was tested with a time lag of
1 year, as can be seen in the model below. The regression equation is as follows:
(3) DON_TOT = β0 + β1 Ti,t-1 + d1-4 SECTOR + d5 SIZE + d6 AGE + d7 TPP +
d8 BIG4 + d9 CASH + d10 SHO + d11 DGENDER + d11 INT + ε
Hypothesis 3 is supported if β1 is positive and significant.
In all regressions, i indicates NGO, t indicates year and e is the error term. The remaining
variables are as explained in Sections 3.2 and 3.3.
31
4. Results
This section elaborates on the results of the archival database analysis. First an outlier analysis
was performed. Thereafter, univariate and bivariate analysis were performed to obtain a better
understanding of the database. Hereby showing the main characteristics of the database and the
relations between the main variables and the control variables. The section concludes by testing
the regression models as presented in Section 3.4 and elaborating on the hypotheses introduced
in Section 2.6.
4.1 Descriptive statistics
According to Field (2013), assumptions are conditions that ensure that what you’re attempting to
analyze works. If these assumptions are violated, then the test statistic and p-value will be
inaccurate and could lead to an inaccurate conclusion. Therefore, the two assumptions being
looked at are normality and multicollinearity (see Section 4.2).
Before the analysis in STATA the assumptions are tested and an outlier analyses was performed.
Normality is the first assumption being analyzed. This assumption means that the data must be
normally distributed. One of the options that Field (2013) gives, to get a normal distribution
from the data, is the removal of outliers. An outlier is a score very different from the rest of the
data, which tend to skew the data distribution (Field, 2013). There is skewness when the
skewness value is > 2 ×Standard deviation of the skewness. The kurtosis is also determined the
same way. When these rules of thumb are violated, then it can be stated that the data is not
normally distributed. Field (2013) also gives the option of performing a Log transformation.
An outlier analysis was performed by analyzing the z score. A z score of ≥ 3.29 or ≤ -3.29 is
considered an outlier. NGOs that provide an outlier for any of the variables is then removed.
After controlling for outliers the initial sample of 162 NGOs is cut down to 150 NGOs.
Furthermore, the following variable has been transformed into normally distributed variable:
DON2=Log(DON_TOT). The central limit theorem states that in large samples (usually n>30)
the estimate will have come from a normal distribution regardless of what the sample data looks
like (Field, 2013). So the assumption of normality is assumed to be the case in this study, which
consists of a sample of 150 NGOs.
32
In Table 7 the descriptive statistics of the sample NGOs is presented, whereby all variables are
untransformed. The descriptive statistics shows the minimum and maximum values, the mean,
the median and the standard deviation for the measurement variables TP, ADMIN, PROG, FR
and DON_TOT and the control variables. The dataset is built up of 150 NGOs, with a total
number of data entries of 900 all together in the years 2010-2015. Over the entire period 27% of
the NGOs has an international parent, 40% are audited by a Big-4 audit firm and 65% have a
male executive director. Figure 2 shows the division among sectors in 2015. The majority of the
NGOs are concerned with international aid (41%) followed by environmental purposes (29%).
Whereas, social welfare and health purposes account for 11% and 19% respectively.
Furthermore, the average organizational age of the examined organizations is 45 years and 6,6%
of the examined NGOs are a member of the Dutch SHO.
Performance is measured by looking at the administrative, program expenditure and fundraising
efficiency. Table 7 shows that for ADMIN the mean is 5,8%. This means that on average the
administrative expenses of NGOs account for around 5,8% of their total costs. In addition,
PROG has a mean value of 82%. This means that NGOs spend 82% of their funds on achieving
organizational goals. The mean value of FR lies at 12%. This performance measurement
indicates the amount of money NGOs have to spend in order to generate funds. Therefore,
throughout the years, the NGOs have to spend €0,12 in order to raise €1 on funds. Donation
levels varies greatly, from a maximum of 148 million to the lowest score of 0, as can be noted in
Table 7. This is also confirmed by the high standard deviation (22,3m), which indicates that the
scores are very much spread out over a large range of values. Figure 3 shows the ordinal scores
for the TP prize in the years 2011-2015. It can be noted that the percentage of NGOs who
scored below the 5.5 decreased gradually every year.
33
Figure 2: Division of NGOs over sectors
Figure 3: Ordinal scores for TP prize
Table 7
Descriptive statistics of all the variables
N Mean Std. deviation Min. value Max. value Median
TP
593
2.554
.8606
1
4
3
DON_TOT 900 13,300,000 22,300,000 0 148,000,000 6,032,555
ADMIN 900 .0581 .0480 0 .5658 .0464
PROG 900 .8173 .1935 0 1.8606 .8389
FR 900 .1195 .1409 0 3.138 .1011
TPP 900 .0089 .0939 0 1 0
AGE 900 45.4867 34.2461 6 224 39
SIZE 900 15.2348 2.1493 7.8148 19.6956 15.5954
CASH 900 .5427 .3036 0 1 0.5658
BIG4 900 .4 .4902 0 1 0
INT_AID 900 .4067 .4915 0 1 0
HEALTH 900 .1867 .3896 0 1 0
ENVIRON 900 .2933 .4555 0 1 0
SOCIAL 900 .1133 .3172 0 1 0
INT 900 .2733 .4459 0 1 0
DGENDER 900 .6467 .4783 0 1 0
SHO 900 .0667 .2496 0 1 0
4.2 Bivariate analysis
Before a correlation matrix is performed, a variance inflation factors (VIF) test is conducted to
test for multilinear relations. The results show that there are no signs of excessive
multicollinearity within the regression models (see Appendix 3). All VIF values are well below 10
and the mean is close to 1.
After controlling for multicollinearity, Spearman-Rho correlations are conducted between the
dependent, independent and the control variable (Table8). The correlation matrix explains the
interrelationship between two variables. The analysis is performed over the six-year period 2010-
2015. The justification for this statistical test is that the main and control variables are measured
on an ordinal or continuous scale. It is also worth noting that Spearman’s correlation is not very
sensitive to outliers (Field, 2013).
As shown in Table 8, TP is significant positive correlated with ADMIN at a 1% level (r=.123).
This positive relation does not support hypothesis 1 that greater accountability improves NGO
performance. This is consistent with the study of Buchneit & Parsons (2006) suggesting that
greater accountability doesn’t necessarily improve NGO performance. PROG and FR are
negatively correlated with TP. Because the correlations between PROG, FR, and TP aren’t
significant, it can’t be concluded that greater accountability doesn’t improve NGO performance.
Hereby, hypothesis 1 is partially supported, because NGOs that show greater accountability
perform less efficient in ADMIN but not in PROG and FR.
Table 8 underlines that over the six-year period all main variables were significantly correlated
with total donations. DON2 is significant negative correlated with ADMIN (r=-.258) at a 1%
level. At the other hand, DON2 is significant and positive correlated with PROG (r=.165) and
FR(r=.123) at a 1% level. These correlations between performance and donations indicate that as
hypothesized, a higher efficiency for ADMIN and PROG is related to higher donation levels.
However, this statement does not hold for FR. What comes in at a surprise is the significant
negative correlation between TP and DON2 at a 1% level. This hints at the fact that higher
accountability may lead to less donations. It could be argued that donors don’t want more
transparency, but more trust. This trust is not or barely determined by the quality of the year
reports. And by allocating energy and resources into being more accountable, donors may feel
that this will go at the expense of pursuing organizational goals.
36
According to the correlation matrix, between ADMIN and PROG, a significant negative
correlation (r=-.39) is found. A similar kind of relation is found between FR and PROG (r=-
.262). This indicates that when a higher amount of resources is spent on organizational
programs, less administration and fundraising costs were made.
Furthermore, AGE and SIZE are significant positive correlated to DON2. This underlines that
organizational size and age, provide some type of goodwill which materialize in higher donation
levels. Also the presence of a big-4 audit firm has a high significant positive influence on
donation levels. That should not be a strange result, taking into account that a big-4 firm
enhances the reliability of the financial information (Yetman & Yetman, 2011). Another
noteworthy correlation is the significant negative correlation between CASH and DON2. The
results are in line with our previous expectation that donors may adjust their donation behavior
if agency costs may arise due to the cash balance. Furthermore, NGOs that are member of SHO
tend to have more donations, as can be noted by the significant positive correlation in Table 8.
NGOs with an international parent have a significant positive association with cash and donation
levels. Moreover, older NGOs seems to be more accountable and raise more funds. Further,
winning the transparency prize seems to lead to less accountability. It could be argued that these
NGOs might become complacent after winning the award.
Table 8
Spearman-Rho matrix for all variables over six-year period
TP DON2 ADMIN PROG FR TPP AGE SIZE CASH BIG4 INT_AID HEALTH ENVIRON SOCIAL INT DGENDER SHO
TP
1
DON2 -.259** 1
ADMIN .123** -.258** 1
PROG -.042 .165** -.39** 1
FR -.044 .123** .219** -.262** 1
TPP -.163* -.080 -.042 -.040 -.017 1
AGE .093* .435** -.047 .060 .083** -.113** 1
SIZE -.124** .854** -.196** .088** .188** -.094* .587** 1
CASH -.109** -.387** .054 -.101* -.17** -.115** -.498** -.514** 1
BIG4 -.178** .588** -.175** .088** .056 -.051 .271** .56** -.254** 1
INT_AID -.194** -.028 -.225** .149** -.33** .047 -.238** -.203** .253** -.066 1
HEALTH .15** .013 .061 -.136* .383** -.047 .066 .151** -.07 .126** -.405** 1
ENVIRON .121** .025 .108** -.049 .033 -.065 .132** .061 -.218** -.041 -.563** -.296** 1
SOCIAL -.054 -.010 .133* .008 .016 .08 .111** .052 -.001 .012 -.289** -.152** -.211** 1
INT -.096* .110** -.050 .021 -.017 -.025 -.146** -.088* .145** .01 .378** -.265** -.123** -.102* 1
DGENDER -.024 .086* .068 .121** .093* -.068 .147** .103* -.109** .1 -.149** -.056 .077 .204** -.002 1
SHO -.078 .357** -.188** .135* -.057 .035 .05 .269** .037 .342** .262** -.132** -.114** -.094* .025** .008 1
**. Correlation is significant at the 0.01 level, based on a two-tailed test.
*. Correlation is significant at the 0.05 level, based on a two-tailed test.
TP: NGO score in transparency prize
4.3 Regression analysis
The regression results will be reported according to the regression equations presented in Section
3.4. Due to the use of time lag, the regression analysis shall be done per hypothesis per year,
whereas the time lag is implemented manually in the models. According to general statistics rules
for dummy variables, one is left out in the model to which the other dummies will be referenced.
In all the models this applies for the dummy variable HEALTH.
4.3.1 Test of hypothesis 1
The first model examines whether a higher level of accountability has an influence on NGO
performance. Performance is examined by looking at the relation of TP on ADMIN, PROG and
FR separately. In the regression estimates, performance proxies are regarded as the dependent
variables, while TP and the other control variables are considered as the independent variables.
To implement the hypothesized time lag, the TP of year before the year stated, will be taken into
account. As accountability scores are only available for the period 2011-2015 and due to the
implementation of a time lag, only four regressions can be run. Namely for the years 2012-2015.
In Table 9, the relationship between ADMIN as proxy for performance and the level of
accountability is shown. According to the regression results, TP is positively related to ADMIN
as this relation holds in all the years (β=.006;.007;.009;.014). Whereas, this result is only
significant at the 5% level in 2015 (β=.014, p=.032). This indicates that a higher accountability
would lead to lower performance level. Therefore, the results do not support the hypothesis that
a higher accountability would lead to a better performance in Dutch NGOs. However, this result
supports the findings of the bivariate analysis (see Section 4.2). Furthermore, only the control
variable SIZE is significant negative related to ADMIN. This result underlines that larger NGOs
are less efficient regarding their administrative expenses.
Table 10 reports the results of the regression analysis using PROG as dependent variable and
proxy for performance. It can be noted that the variable TPP has a significant positive
relationship with PROG in 2012 (β=.250, p=.030). This means that NGOs who won the
transparency prize in 2012 perform better. Furthermore, the variables SIZE and DGENDER
show a significant positive relationship with PROG in 2015. These results underline that larger
organization and with a male executive perform better.
39
The last regression analyzes the relationship of TP on the dependent variable FR (Table 11).
Only the variables BIG4, INT_AID, ENVIRON and SOCIAL were significant, namely in the
year 2012. The control variable BIG4 was significant on a 5% significance level (β=-.043,
p=.038). This result means that having a big-4 audit firm auditing your financial information, is
associated with a higher performance level.
In summary the above results show that there is no significant effect of TP on NGO
performance. So it is not proven that the more accountable an NGO is, the better they perform.
Therefore, hypothesis 1 is not supported.
Table 9: Regression estimates of accountability on performance (ADMIN)
ADMIN 2012 2013 2014 2015
TP
.006
(.910)
.007
(1.450)
.009
(1.540)
.014*
(2.170) TPP -.044
(-.990) .002
(.060) .002
(.030) .017
(.350) AGE .000
(1.080) -.00001 (-.100)
.000 (1.630)
.000 (.520)
SIZE -.012** (-3.040)
.000 (.140)
-.001 (-.410)
-.002 (-.500)
CASH .015 (.670)
.034 (1.900)
.012 (.690)
.018 (.940)
BIG4 .012 (.840)
-.004 (-.360)
-.013 (-1.190)
-.011 (-1.030)
INT_AID -.01 (-.590)
-.028 (-1.120)
-.018 (-1.150)
-.021 (-1.300)
ENVIRON -.008 (-.500)
-.009 (-.540)
-.004 (-.290)
.003 (.160)
SOCIAL .008 (.400)
-.020 (-1.120)
.011 (.660)
.019 (1.030)
INT -.007 (-.330)
.009 (.940)
-.002 (-.190)
-.00004 (-.0001)
DGENDER -.004 (-.200)
.007 (.760)
.004 (.420)
.012 (-1.300)
SHO -.005 (-.2)
-.014 (-.730)
-.007 (-.380)
-.013 (-.730)
N 129 126 114 114
Adjusted R .062 .024* .044* .087
**, * indicate statistical significance at 1 and 5% level respectively (two tailed; t-values below the regression coefficients in parentheses) Model 1.1: ADMIN = β0 + β1 Ti,t-1 + d1-4 SECTOR + d5 SIZE + d6 AGE + d7 TPP + d8 BIG4 + d9 CASH + d10 SHO + d11 DGENDER + d11 INT + ε
Table 10: Regression estimates of accountability on performance (PROG)
PROG 2012 2013 2014 2015
TP
-.016
(-1.050)
.017
(1.020)
.040
(-1.130)
.054
(1.930) TPP .250*
(2.190) .009
(.070) .020
(1.300) .134
(.640) AGE -.001
(-1.780) -.0003 (-.680)
-.0007 (-1.130)
-.0006 (-1.010)
SIZE .014 (1.390)
.007 (.670)
.013 (.930)
.047* (3.480)
CASH -.111 (-1.970)
-.024 (.730)
-.002 (-.020)
.044 (.520)
BIG4 -.025 (-.700)
.027 (.730)
.015 (.300)
.013 (.280)
INT_AID .085 (1.950)
.097 (1.560)
.066 (.900)
.067 (.920)
ENVIRON .050 (1.210)
.058 (.950)
-.021 (-.290)
.016 (.230)
SOCIAL .019 (.360)
.053 (.830)
.136 (1.670)
.078 (.950)
INT .006 (.180)
-.063 (-1.730)
-.007 (-.150)
-.031 (-.680)
DGENDER .046 (1.640)
.056 (1.760)
-.005 (-.130)
.115* (-.680)
SHO .059 (.94)
.040 (-1.730)
-.022 (-.270)
-.045 (-.580)
N 129 126 114 114
Adjusted R .104 .045* .022* .164
**, * indicate statistical significance at 1 and 5% level respectively (two tailed; t-values below the regression coefficients in parentheses)
Model 1.2: PROG = β0 + β1 Ti,t-1 + d1-4 SECTOR + d5 SIZE + d6 AGE + d7 TPP + d8 BIG4 + d9 CASH + d10 SHO + d11 DGENDER + d11 INT + ε
42
Table 11: Regression estimates of accountability on performance (FR)
FR 2012 2013 2014 2015
TP
-.011
(-1.250)
-.005
-(.500)
-.009
(-.890)
.002
(.200) TPP .014
(.220) .067
(.900) .200
(1.300) .045
(.530) AGE -.0003
(-.104) -.0006
(-1.800) -.0003
(-1.650) -.0002
(-1.060) SIZE .010
(1.620) .015*
(2.610) .007
(1.530) .005
(.900) CASH 4.6E-05
(.000) -.009
(-.240) -.043
(-1.460) -.051
(-1.490) BIG4 -.043*
(-2.100) -.063*
(-2.920) -.035
(-1.910) -.018
(-.970) INT_AID -.156**
(-6.130) -.042
(-1.170) -.046
(-1.740) -.045
(-1.540) ENVIRON -.107**
(-4.460) .004
(.120) -.016
(-.620) -.022
(-.760) SOCIAL -.141**
(-4.650) .115
(3.040) -.051
(-1.750) -.036
(-1.110) INT .024
(1.200) .024
(1.110) .017
(1.020) .015
(.830) DGENDER .006
(.340) -.003
(-.160) .006
(.360) .019
(1.160) SHO .022
(.60) .028
(.700) .015
(.510) -.021
(-.680)
N 129 126 114 114 Adjusted R .237 .263 .073 .077
**, * indicate statistical significance at 1 and 5% level respectively (two tailed; t-values below the regression coefficients in parentheses)
Model 1.3: FR = β0 + β1 Ti,t-1 + d1-4 SECTOR + d5 SIZE + d6 AGE + d7 TPP + d8 BIG4 + d9 CASH + d10 SHO + d11 DGENDER + d11 INT + ε
4.3.2 Test of hypothesis 2
In this section hypothesis 2 is tested, which examines the effects of performance on donation
levels. Performance is examined by looking at the relation of ADMIN, PROG and FR
separately, on donation levels (DON2). In the regression estimates, total donations is regarded as
the dependent variable, while the performance proxies and the other control variables are
considered as the independent variables. The regression models as presented in Section 3.4 are
tested for the years 2010-2015.
As hypothesized, there is a significant negative relation between ADMIN and donations in 2012
at the 5% significance level (β=-5.234, p=.000). This means that the more efficient an NGO
perform, the higher the donation levels. But in 2010 the results show the contrary, namely a
significant positive relation between ADMIN and DON2 (β=12.314, p=.022) at the 5%
significance level (Table 12). As such, the results do not support the hypothesis that a better
performance leads to higher donation levels. The significant positive coefficient for INT in 2012
(β=.679, p=.000) indicates that the presence of an international parent leads to a higher donation
level. This relationship also holds for model 2.2 in 2015 and model 2.3 in 2013.
Table 13 shows the results for model 2.2. As expected, program expenditure is positively related
with donation levels. Showing a significant relation for all the years. Therefore, the results
support the hypothesis that a better performance would lead to higher donation levels.
The last model 2.3 will investigate the influence of FR on donation levels. In line with the
findings of the bivariate analysis, the results show a positive coefficient for the relationship
between FR and DON2. For 2010-2014 this relationship is significant at the 1% significance
level, as can be noted in Table 14. The significant positive coefficient for DGENDER in model
2.3 indicates that the presence of a male executive leads to higher donation levels. Whereas the
significant positive coefficient for the sector INT_AID and ENVIRON indicates that NGOs in
these sectors generate more donations.
Moreover, the variable SIZE has a significant positive coefficient throughout all of the models.
This result is in line with prior research (Bekker & Wiepking, 2011; Parsons, 2007) and indicates
that larger NGOs generate more donations.
44
In summary the above results show that there is a significant positive effect of PROG on the
donation levels. But what is previously hypothesized does not hold for the performance
indicators ADMIN and FR. Therefore, hypothesis 2 is partially supported.
Table 12: Regression estimates of performance on donation levels (ADMIN)
DON2 2010 2011 2012 2013 2014 2015
ADMIN 12.314* (2.310)
6.235 (1.290)
-5.234** (-4.380)
-1.344 (-.490)
-.580 (-.200)
5.815 (1.600)
TPP -.667 (-.340)
-1.102 (-.480)
-.834 (-1.350)
.709 (.610)
.605 (.520)
1.065 (.530)
AGE .0007 (.100)
-.004 (-.650)
-.002 (-.790)
-.006 (-1.410)
-.006 (-1.320)
-.010 (-1.840)
SIZE .994** (7.020)
.894** (7.660)
.692** (15.650)
.965** (11.830)
.968** (11.690)
1.200** (11.850)
CASH 1.016 (1.130)
(.944) (1.230)
.121 (.420)
.149 (.280)
.052 (.100)
.327 (.500)
BIG4 .412 (.700)
.312 (.690)
.301 (1.800)
.169 (.540)
.076 (.230)
-.119 (-.310)
INT_AID .472 (.700)
.856 (1.480)
-.026 (-.120)
.687 (1.720)
.750 (1.860)
.537 (1.110)
ENVIRON .007 (.010)
.037 (.070)
.069 (.320)
.605 (1.540)
.594 (1.500)
.457 (.970)
SOCIAL .257 (.310)
.268 (.370)
-.131 (-.490)
.225 (.450)
.282 (.560)
.245 (.410)
INT .929 (1.730)
.272 (.580)
.679** (3.930)
.398 (1.240)
.412 (1270)
.484 (1.240)
DGENDER .581 (1.250)
.195 (.490)
.135 (.910)
.569* (2.070)
.514 (1.830)
.742* (2.200)
SHO .112 (.110)
.113 (.130)
.133 (.420)
.021 (.040)
.073 (.120)
.181 (.250)
N 150 150 150 150 150 150 Adjusted R .358 .353 .793 .626 .624 .609
**, * indicate statistical significance at 1 and 5% level respectively (two tailed; t-values below the regression coefficients in parentheses)
Model 2.1: DON2 = β0 + β2 ADMIN + d1-4 SECTOR + d5 SIZE + d6 AGE + d7 TPP + d8 BIG4 + d9 CASH + d10 SHO + d11 DGENDER +
d11INT + ε
46
Table 13: Regression estimates of performance on donation levels (PROG)
DON2 2010 2011 2012 2013 2014 2015
PROG 7.900** (10.09)
6.780** (8.400)
1.550** (3.260)
6.159** (9.350)
3.765** (6.370)
5.127** (7.730)
TPP -1.270 (-.850)
.904 (.470)
-1.011 (-1.570)
1.116 (1.230)
1.202 (1.310)
.732 (.430)
AGE .002 (.510)
-.008 (-.152)
-.001 (-.510)
-.003 (-.730)
-.003 (-.840)
-.007 (-1.580)
SIZE .893** (8.200)
.886** (9.350)
.721** (16.180)
.982** (15.440)
.961** (13.270)
1.114** (13.040)
CASH 1.127 (1.630)
(.734) (1.170)
.251 (.840)
.329 (.780)
.093 (.200)
.300 (.550)
BIG4 .332 (.830)
-.132 (-.360)
.332 (1.940)
-.020 (-.080)
.001 (.000)
-.250 (-.780)
INT_AID -.503 (-.960)
.400 (.850)
-.060 (-.270)
.014 (.040)
.202 (.560)
-157 (-.380)
ENVIRON -.207 (-.410)
.240 (.520)
.018 (.080)
.031 (.100)
.327 (.930)
.224 (.560)
SOCIAL -.173 (-.270)
.212 (.360)
-.173 (-.630)
-.234 (-.590)
-.044 (-.100)
.087 (.170)
INT .692 (1.670)
.467 (.122)
.688** (3.870)
.724 (2.850)
.515 (1.810)
.891** (2.690)
DGENDER -.186 (-.510)
.297 (.910)
.090 (.580)
.256** (1.170)
.438 (1.780)
.283 (.980)
SHO .211 (.270)
-.051 (-.070)
.070 (.220)
-.331 (-.710)
.126 (.240)
.156 (.250)
N 150 150 150 150 150 150 Adjusted R .618 .568 .782 .770 .710 .723
**, * indicate statistical significance at 1 and 5% level respectively (two tailed; t-values below the regression coefficients in parentheses)
Model 2.2: DON2 = β0 + β3 PROG + d1-4 SECTOR + d5 SIZE + d6 AGE + d7 TPP + d8 BIG4 + d9 CASH + d10 SHO + d11 DGENDER +
d11INT + ε
47
Table 14: Regression estimates of performance on donation levels (FR)
DON2 2010 2011 2012 2013 2014 2015
FR 7.167** (2.550)
6.543** (2.850)
1.135** (-4.040)
2.307 (1.590)
3.751* (2.320)
2.168 (1.310)
TPP -.648 (-.330)
-.956 (-.420)
-.605 (-.970)
.497 (.430)
.510 (.480)
1.016 (.500)
AGE .0002 (.230)
-.002 (-.320)
-.003 (-1.370)
-.006 (-1.230)
-.005 (-.980)
-.008 (-1.530)
SIZE .873** (6.050)
.795** (6.800)
.778** (17.350)
.936** (11.260)
.923** (11.110)
1.163** (11.350)
CASH 1.017 (1.140)
(.778) (1.040)
.219 (.080)
.158 (.290)
.078 (.150)
.319 (.490)
BIG4 .533 (1.020)
.440 (1.000)
.209 (1.220)
.292 (.920)
.248 (.770)
-.078 (-.200)
INT_AID 1.031 (1.400)
1.543 (2.480)
.287 (1.300)
1.048* (2.350)
1.236** (2.780)
.732 (1.370)
ENVIRON .532 (.780)
.566 (.980)
.226 (1.050)
.846* (2.020)
.933* (1.225)
.727 (1.440)
SOCIAL 1.052 (1.200)
.956 (1.320)
.056 (.210)
.534 (1.000)
.689 (1.320)
.529 (.830)
INT .721 (1.340)
.089 (.190)
.586** (3.320)
.323 (1.000)
.313 (.980)
.474 (1.210)
DGENDER .524 (1.140)
.148 (.380)
.119 (.800)
.563* (2.060)
.537 (1.950)
.765* (2.270)
SHO .052 (.050)
.128 (.150)
.166 (.520)
.018 (.030)
.079 (.140)
.154 (.210)
N 150 150 150 150 150 150 Adjusted R .363 .384 .790 .632 .638 .607
**, * indicate statistical significance at 1 and 5% level respectively (two tailed; t-values below the regression coefficients in parentheses)
Model 2.3: DON2 = β0 + β4 FR + d1-4 SECTOR + d5 SIZE + d6 AGE + d7 TPP + d8 BIG4 + d9 CASH + d10 SHO + d11 DGENDER +
d11INT + ε
4.3.3 Test of hypothesis 3
In this section hypothesis 3 is tested, which examines whether higher accountability is positively
related with donation levels. In the regression estimates, donation levels is regarded as the
dependent variable, while the accountability proxy and the other control variables are considered
as the independent variables. To implement the hypothesized time lag, the TP of year before the
year stated, will be taken into account. As accountability scores are only available for the period
2011-2015 and the implementation of a time lag, only four regressions can be run. Namely for
the years 2012-2015.
Hypothesis 3 is supported if β1 is positive and significant. As can be noted in Table 16, in 2014
(β=.107) and 2015 (β=.126) there is a positive coefficient for accountability. However, the results
are not significant at the 5% significance level. Therefore, the results don’t explicitly suggest that
higher accountability leads to more donations. Thus, providing no support for hypothesis 3.
Furthermore, the control variable SIZE shows a significant positive relation with DON2. This
indicates that the larger the NGO, the higher the donation level. Other variables that show a
significant positive relation with DON2 are DGENDER, INT_AID and INT. For an overview
of the hypotheses results, see Table 15.
Table 15: Hypotheses conspectus
2010 2011 2012 2013 2014 2015
H1 n.a. n.a. Rejected Rejected Rejected Rejected
H2 Confirmed Confirmed Confirmed Confirmed Confirmed Confirmed
H3 n.a. n.a. Rejected Rejected Rejected Rejected
Table 16: Regression estimates of accountability on donation levels
DON2 2012 2013 2014 2015
TP
-.240** (-2.950)
-.233
(1.360)
.107
(.470)
.126
(.430) TPP -.906
(-1.520) 1.471
(1.150) 1.250 (.900)
.968 (.440)
AGE -.005* (-2.020)
-.007 (-1.300)
-.007 (-1.330)
-.009 (-1.370)
SIZE .843** (15.960)
1.040** (10.490)
1.156** (10.380)
1.500** (10.520)
CASH -.260 (-.880)
.020 (.030)
.192 (.280)
1.240 (1.400)
BIG4 -.220 (-1.180)
.220 (.580)
-.214 (-.510)
-.376 (-.770)
INT_AID -.131 (-.580)
.954* (2.020)
1.032 (1.970)
.612 (.950)
ENVIRON .076 (.350)
.690 (1.520)
.847 (1.650)
.770 (1.250)
SOCIAL -.271 (-1.000)
.037 (1.520)
.395 (.590)
.344 (.400)
INT .677* (1.780)
.358 (.970)
.554 (1.420)
.895 (1.880)
DGENDER .054 (.370)
.675* (2.100)
.435 (1.220)
.847 (1.970)
SHO .483 (1.480)
.003 (.000)
.051 (.070)
-.195 (-.240)
N 129 126 114 114
Adjusted R .825 .629 .645 .628
**, * indicate statistical significance at 1 and 5% level respectively (two tailed; t-values below the regression coefficients in parentheses) Model 3: DON2 = β0 + β1 Ti,t-1 + d1-4 SECTOR + d5 SIZE + d6 AGE + d7 TPP + d8 BIG4 + d9 CASH + d10 SHO + d11 DGENDER + d11 INT + ε
5. Conclusion
In this section, a summary of the study is presented, whereby the results are further explained.
Finally, implications and limitations of the study will be provided.
5.1 Summary
This paper investigated the relation between accountability, performance and donation levels.
To answer the research question, the effect of accountability on performance and donation levels
were assessed. Moreover, it evaluated the effect of performance on donation levels. This paper
complements transparency prize data (PWC archive) with financial information (CBF database).
The sample for this study is 150 CBF accredited NGOs in the years 2010-2015, who participated
in the transparency prize in the years 2011-2015.
A necessary input for the survival of an NGO is achieving a balance between the interests of all
stakeholders and legitimizing their position in society (Deegan & Blomquist, 2006). Therefore,
the board needs to be accountable to the various stakeholders and communicate whether
management performance is in line with their expectations. Accordingly, we hypothesize that
higher accountability leads to a better NGO performance. However, the results on the relation
between accountability and performance do not support the hypothesis. A positive instead of
negative sign of the coefficient was found for the performance indicator ADMIN. Furthermore,
the interaction between accountability and the performance indicators PROG and FR were
insignificant. Thus H1 is not supported. This is consistent with the results found by Buchneit &
Parsons (2006), who states that a greater accountability doesn’t necessarily improve NGO
performance.
For the second hypothesis, the influence of performance on donation levels, regression analysis
found a significant positive interaction only when using PROG as a proxy for performance. This
is consistent with the research of ING (2016), which states that donors base their donation
decisions on the reputation of the NGO. Therefore, when a NGO is well known for spending a
high percentage of the revenue on achieving organizational goals, it could lead to higher
donation levels. At the other hand, when using FR as the performance indicator, a significant
positive relation has been found instead of a negative sign of the coefficient. It could be argued,
51
that donors don’t base their donation behavior on whether NGOs generate funds efficiently.
The more an NGO spend on fundraising activities, the better the brand awareness which leads
to more donations.
Whereas Marudas & Jacobs (2009) found a negative relation between ADMIN and donation
levels, the results of this study only found such a significant interaction in 2012. A reason that
might help to explain this result is the sample. Research by Jacobs & Marudas (2009) employ
data from the US. The donation behavior of the donors may not be the same as in the Dutch
sector due to cultural differences which can influence donor behavior. According to the results,
it can be stated that donors in the Dutch NGO sector pay less attention whether an NGO
perform efficiently regarding administration and fundraising costs. This behavior could be
explained by the fact that the CBF prescribes his accredited members (with revenues > 2m) that
fundraising costs need to be a maximum of 25% of the fund revenues. Therefore, a donor
choosing to donate to an CBF accredited NGO may pay less attention to these performance
measures, as they assume that the organizations behavior is in line with the governance code of
the CBF. Thus, as discussed above, H2 is partially supported.
The research of ING (2016) states that an increase in the level of transparency trough a report or
other forms of media coverage, would enhance donors trust and lead to a stronger relation with
the donors. Accordingly, we hypothesized that higher accountability is positively related with
donation levels (H3). The regression analysis does not support the premises.
According to Gent et al. (2014), to attract funding, NGOs have an incentive to focus their
efforts on achieving short term quantifiable accomplishments. Herein lies the reputation trap, as
NGOs are focused on continuous production of tangible results to maintain their reputation and
survive financially, instead of focusing their efforts into more durable outcomes. This short term
focus can cause mission drift and is one of the negative externality of NGO accountability
mechanisms (Gent et al., 2014). As mentioned above, H3 found no significant positive relation
between accountability and donation levels. Thus, the NGO-donor relationship is not improved
by strictly improvements in transparency and accountability (Gent et al., 2014), as the trend has
been in recent years.
52
Concluding, the Dutch NGO sector lies in a reputation trap for no reason. They put their efforts
in improving accountability, which could lead to agency problems as NGOs may pursue short
term successes in order to communicate it through their improved yearly reports and gain
reputation. But donors don’t seem to be interested in more transparency. It could be argued that
they want more trust that their donations are converted into quality investments towards
achieving the underlying mission. But without reliable transparent information, there is no trust
(ING, 2016). Therefore, moving forward, the NGO sector must find the optimal balance
between quality communication and quality impact investment: walk the talk (De Dikke Blauwe,
2017).
5.2 Implications
The results of this paper have several implications. This paper fills the gap in prior studies by
combining the interactions between accountability, performance and donation levels. The found
significant relationship between program funding (PROG) and performance, indicate that
donors pay more attention to the efficient usage of program funds than to administration and
fundraising costs. To enhance donation levels, NGOs should therefore focus their policies on
program funding rather than administration and fundraising costs.
Furthermore, the results question the accountability practices by Dutch NGOs. Given the
insignificant relation found between accountability and donation levels, it could be questioned if
NGOs are not better off by putting their efforts into durable organizational outcomes, instead of
satisfying certain bureaucratic demands. It is recommended for NGOs to implement policies
that can balance the extent of accountability and performance measurement, without drifting
away from the mission. The results also have implications for regulators and standard setters. If
they are aware of the level of association, it could help to develop a framework to further
improve the NGO-donor relationship. The NGO-donor relationship needs to be improved in
order the mitigate the effects of the reputation trap (Gent et al., 2014).
5.3 Limitations and recommendations for future research
There are also limitations in this study that should be acknowledged and could be helpful to
guide future research. A limitation of this study is the proxies used for performance
53
measurement. As the non-profit sector lacks a commonly accepted indicator for performance
(Miller, 2002), this study has used three different proxies for performance. The aspect that has
been looked is efficiency. At the other hand, NGOs should also be evaluated on their impact and
whether expected organizational goals are reached. As there is no single variable that can capture
all aspects of NGO performance, a limitation has to be recognized. That said, future research
should try to gather a proxy that captures both efficiency and effectiveness when measuring
NGO performance. This could improve the results on this field of study.
A second limitation is the proxy used to measure accountability. As transparency is the literal
value of accountability (Koppel, 2005), the transparency prize served as the measure for
accountability in this study. It is possible that this proxy does not capture all aspects of
accountability as shown in Table 2. This limitation is therefore also a recommendation for future
research to try to gather a proxy that would capture a much broader aspect of accountability.
A final limitation is that limiting the data sample to NGOs that have a CBF certification mark
introduces bias in the sample. It is recommended for future research to expand their data sample
to non CBF certified NGOs as this could influence the results.
What could also be done in future research is expand the sample into an international
perspective and not only focus on Dutch NGOs. The regression results may reverse by taking a
more global approach, as different cultures might influence the donation behavior. Research on
these issues should lead to a more complete understanding on the relation between
accountability, performance and donation levels in the NGO sector.
54
References
Aggarwal, R., Evans, M., & Nanda, D. (2011). Non profit boards: Size, performance and
managerial incentives. Journal of Accounting and Economics, article in press.
Ahrens, T., & Chapman, C. (2006). Doing qualitative field research in management accounting:
positioning data to contribute to theory. Accounting, Organizations and Society, 31(8), 819-
841.
Beamon, B., & Balcik, B. (2008). Performance measurement in humanistairian relief chains.
International journal of public sector management, 1, 4-25.
Bekkers, R., & Wiepking, P. (2011). A literature review of empirical studies of philanthropy:
mechanisms that drive charitable giving. . Nonprofit and Voluntary Sector Quarterly, 40(5),
924-973.
Biddle, G., Hillary, G., & Verdi, R. (2009). How does financial reporting quality relate to
investment efficiency?. Journal of Accounting and Economics, 48, 112-131.
Brandsen, T., & Pape, U. (2015). The Netherlands: The paradox of Government-Nonprofit
Partnerships. Voluntas, 26, 2267-2282.
Burger, R., & Owens, T. (2010). Promoting transparency in the NGO sector: Examining the
availability and Reliability of self reported data. World Development, 28(9), 1263-1277.
Cameron, W. (2004). Public accountability: Effectiveness, equity, ethics. National Council of the Institute
of Public Administration.
CBF. (2016). Jaarverslag CBF 2016. Retrieved July 19, 2017, from http://www.cbf.nl/over-het-
cbf/jaarverslagen/
CBF. (2017). Overzicht baten en lasten. Retrieved July 18, 2017, from www.cbf.nl/sector-brede-
informatie/baten-en-lasten
Cho, C., Guidry, R., Hageman, A., & Patten, D. (2012). Do actions speak louder than words? An
empirical investigation of corporate environmental reputation. Accounting, Organization and
Society, 37, 14-25.
De Dikke Blauwe. (2017, May 17). news. Retrieved June 1, 2017, from
https://www.dedikkeblauwe.nl/news/minder-transparantie-en-meer-vertrouwen-graag
55
Deegan, C., & Blomquist, C. (2006). Stakeholder influence on corporate reporting: An
exploration of the interaction between WWF-Australia and the Australian minerals
industry. Accounting, Organizations and Society, 31, 343-372.
Ebrahim, A. (2003a). Making sense of accountability: Conceptual perspectives for Northern and
Southern nonprofits . Nonprofit Management and Leadership, 14(2), 191-212.
Ebrahim, A. (2003b). Accountability in practice: Mechanisms for NGO’s. World Development, 31,
813-829.
Ebrahim, A. (2005). Accountability myopia: Losing sight of organizational learning . Nonprofit and
Voluntary Sector Quarterly, 34(1), 56-87.
Ebrahim, A. (2009). Placing the normative logics of accountability in ‘thick’ perspective. American
Behavioral Scientist, 52(6), 885-904.
Edwards, M. (2000). NGO Rights and Responsibilities: A New Deal for Global Governance. London:
The Foreign Policy Centre/NCVO.
Edwards, M., & Hulme, D. (1996b). Too Close for Comfort? The Impact of Official Aid on
Nongovernmental Organizations. World Development, 24(6), 961-973.
Field, A. (2013). Discovering Statistics using IBM SPSS Statistics. London: Sage Publications.
Frumkin, P., & Keating, E. (2010). The price of doing good: Executive compensation in
nonprofit organizations. Policy and Society, 29, 269-282.
Frumkin, P., & Kim, M. T. (2001). Strategic Positioning and the Financing of Nonprofit
Organisations: Is Efficiency Rewarded in the Contributions Marketplace. Public
administration review , 61, 266-275.
Gent, S., Crescenzi, M., Menninga, E., & Reid, L. (2014). The reputation trap of NGO accountability.
University of North Carolina at Chapel Hill, Department of Political Science.
Gray, R., Bebbington, J., & Collison, D. (2006). NGOs, civil society and accountability: making
the people accountable to capital. Accounting, Auditing & Accountability journal, 19(3), 319-
348.
Greenlee, J. S., & Brown, K. L. (1999). The Impact of Accounting Information on Contributions
to Charitable Organizations. Research in Accounting Regulation, 13, 111-125.
Hansmann, B. (1980, April). The role of Nonprofit Enterprise . The Yale Law journal, 89(5), 835-
901.
56
Hyndman, N., & McDonnell, P. (2009). Governance and Charities: An exploration of key
themes and the development of a research agenda. Financial Accountability & Management ,
25(1), 0267-4424.
ING. (2016, November). Van donatie naar relatie: naar een totale klantenbeleving voor goede doelen.
Retrieved January 15, 2017, from https://www.ing.nl/zakelijk/kennis-over-de-
economie/uw-sector/publieke-sector/van-donatie-naar-relatie.html
Jacobs, F., & Marudas, N. (2009). The combined effect of donation price and administrative
inefficiency on donations to US non-profit organizations. Financial Accountability &
Management, 25(1), 33-53.
Kähler, J., & Sargeant, A. (2002). The size effect in the administration costs of charities. . The
European Accounting Review, 11(2), 215-243.
Kaplan, R. (2001). Performance measurement and management in nonprofit organizations.
Nonprofit Management & Leadership, 11(3), 353-370.
Keller, G. (2005). Statistics for Management and Economics (7th ed.). Mason, Ohio, USA: Thomson
Higher Education.
Koppell, J. (2005). Pathalogies of Accountability: ICANN and the Challenge of 'Multiple
Accountabilities Disorder'. Public Administration Review, 94-108.
Krishnan, R., Yetman, M. H., & Yetman, R. (2006). Expense misreporting in nonprofit
organizations. Accounting Review, 81(2), 399-420.
Marudas, N., & Jacobs, F. (2006). Effects of Nonprofit Organizational Wealth on Donations:
Evidence from Recent Data on the NonProfit Times 100. Zeitschrift fur Offentliche und
Gemeinwirtschaftliche Unternehmen, 34, 74-88.
Miller, J. (2002). The board as a monitor of organizational activity. Nonprofit Management &
Leadership, 12(4), 429-450.
Mitchell, R., Bradley, R., & Wood, D. (1997). Toward a Theory of Stakeholder identification and
salience: Defining the principle of who and what really counts. The Academy of Management
Review, 22(4), 853-886.
O'Dwyer, B., & Boomsma, R. (2015). The co-construction of NGO accountability: aligning
imposed and felt accountability in NGO-funder accountability relations. Accounting,
Auditing and Accountability Journal, 28, 36-68.
57
O'Dwyer, B., & Unerman, J. (2007). From functional to social accountability: Transforming the
accountability relationship between funders and non-governmental development
rganisations. Accounting, Auditing and Accountability Journal, 20(3), 446-471.
O'Dwyer, B., & Unerman, J. (2008). The paradox of greater NGO accountability: A case study
of Amnesty Ireland. Accounting Organizations and Society, 33, 801-824.
OECD. (2011). The Development Co-operation Report 2011. Retrieved January 14, 2017, from
www.oecd. org/document/10/0,3746,en_2649_33721_44774218_1_1_1_1,00.html
OECD. (2012). Efforts towards a common standard for transparency. Retrieved January 14, 2017, from
http://www.oecd.org/dachlm/DACHLMTechDocTransparency.pdf
O'leary, S. (2017). Grassroots accountability promises in rights-based approaches to
development: The role of transformative monitoring and evaluationin NGOs. Accounting,
Organizations and Society, 1-21.
Parsons, L. (2007). The impact of financial information and voluntary disclosures on
contributions to not-for-profit organizations. Behavioral Research in Accounting, 19(1), 179-
196.
Posnett, J., & Sandler, T. (1989). Demand for Charity Donations in Private Nonprofit Markets.
Journal of Public Economics, 40, 187-200.
Ruben, R., & Schulpen, L. (2009). How to allocate public funding to nongovernmental
development organizations: a critical assessment of the Dutch co-financing system.
Nonprofit and Voluntary Sector Quarterly, 38(2), 287-306.
Samenwerkende Hulporganisatie. (2017). over SHO. Retrieved March 22, 2017, from
www.giro555.nl
Schmitz, H., Raggo, P., & Bruno-van Vijfeijken, T. (2012). Accountability of transnational
NGOs: aspirations vs practice. Nonprofit and Voluntary Sector Quarterly, 41(6), 1175-1194.
Sell, S., & Prakash, A. (2004). Using Ideas Strategically: The contest between Business and NGO
networks in intellectual property rights. International Studies Quarterly, 48, 143-175.
Shankman, N. (1999). Reframing the debate between agency and stakeholder theories of the
firm. . Journal of Business Ethics, 19, 319-334.
Steinberg, R. (2006). Economic theories of nonprofit organizations. (Yale University Press, New Haven,
CT). The Nonprofit Sector Research Handbook.
58
Tinkelman, D., & Mankaney, K. (2007). When is Administrative Efficiency Associated with
Charitable Donations. Nonprofit and Voluntary Sector Quarterly, 36, 41-64.
Transparant Prijs. (2015). Retrieved January 15, 2017, from
http://www.pwc.nl/nl/transparantprijs/archief/publicaties.html
Trussel, J., & Parsons, L. (2008). Financial reporting factors affecting donations to charitable
organizations. Advances in Accounting, 23(1), 263-285.
Tulder, R. v., & Zwart, A. v. (2006). International business-society management: Linking corporate
responsibility and globalization. London: Routledge.
Unerman, J., & O'Dwyer, B. (2006a). On James Bond and the importance of NGO
accountability. Accounting, Auditing and Accountability Journal, 19, 305-318.
Unerman, J., & O'Dwyer, B. (2006b). Theorising accountability for NGO advocacy. Accounting,
Auditing and Accountability Journal, 19(3), 349–376. Accounting, Auditing and Accountability
Journal, 19(3), 349-376.
Weisbrod, B., & Dominguez, N. (1986). Demand for collective goods in private nonprofit
markets: can fundraising expenditures help overcome free-rider behavior? Journal of Public
Economics, 30(1), 83-96.
Yetman, R., & Yetman, R. (2011). The Effects of Governance on the Accuracy of Charitable
Expenses Reported by Nonprofit Organizations. Contemporary Accounting Research, 738-
767.
59
Appendix
Appendix 1
List of abbreviations
CBF Centraal Bureau Fondsenwerving (Dutch institute for the monitoring fundraising
organizations)
ER Erkenningsregeling (CBF certification mark as of 2016, replacing the existing CBF
certification mark, RfB keur and Keurmerk Goede Doelen)
FI Nederland Filantropieland (Dutch institute for philanthropic knowledge exchange and
activities)
GDN Goede Doelen Nederland (Dutch branch organization)
NGO Non-Governmental Organization
NPO Nonprofit Organization
PWC PricewaterhouseCoopers (Big-4 audit firm)
VIF Variance inflation factors
60
Appendix 2
Transparency prize criteria
61
Appendix 3
Multicollinearity
H1/H3 H2
VIF 1/VIF VIF 1/VIF
INT_AID 2,63 0,38 INT_AID 3,24 0,309
SIZE 2,18 0,46 ENVIRON 2,68 0,373
ENVIRON 2,01 0,5 SOCIAL 2,43 0,412
CASH 1,61 0,623 SIZE 1,89 0,528
BIG4 1,6 0,626 CASH 1,57 0,636
SOCIAL 1,52 0,66 AGE 1,45 0,69
AGE 1,41 0,707 BIG4 1,44 0,695
SHO 1,37 0,727 SHO 1,33 0,754
INT 1,25 0,8 INT 1,27 0,789
TP 1,23 0,816 ADMIN 1,13 0,888
DGENDER 1,09 0,92 FR 1,12 0,89
TPP 1,08 0,93 PROG 1,11 0,904
DGENDER 1,07 0,931
MEAN VIF 1,58 TPP 1,03 0,976
MEAN VIF 1,62
*HEALTH = omitted